AI doesn't get tired at 9pm on a Thursday
How always-on conversation handling reshaped our booking curve in the first 90 days.
When we instrumented reply latency properly in April, the first thing the data made obvious was how much pipeline was leaking to evenings and weekends. A meaningful fraction of developer replies land outside business hours - they are running their own businesses, they reply when they reply. Human BDR teams cannot economically cover that. The model can, and the booking curve reshapes accordingly.
Where the replies actually land
Across April-June 2025 on Capital Edge accounts, 41% of developer-initiated messages landed outside 9-6 weekday hours. The largest single bucket was 8pm-11pm Sunday through Thursday - developers catching up on inbox at the end of the day.
A human BDR team handles those Monday morning. A model handles them in ninety seconds. That single shift moved booked-call rate by 22% across the cohort, holding everything else constant.
Why latency matters this much
Developer intent on inbound is perishable. A prospect who replied at 9pm because they were thinking about a deal is, at 9am the next morning, thinking about three other things. The model catches them in the window where the deal is still top of mind. The human BDR mostly catches the residue.
It is not that the messages the model sends at 9pm are better than the ones the BDR would send at 9am. They are not. They just arrive at the right time.
The Sunday afternoon effect
Sunday afternoon is now our second-highest booking window of the week. Three years ago that would have been unimaginable - no developer was being booked into a sales call on Sunday by a human BDR. The model has no opinion on the day of the week, replies normally, and the developer books because they are thinking about the week ahead.
Always-on is not a small advantage. It is a structural one.
The 9pm Thursday problem is one of the cleanest illustrations of why this technology is not just 'faster BDRs'. It is a different shape of coverage. Once a broker sees the booking curve fill in on evenings and weekends, going back to business-hours-only feels like turning off a meaningful chunk of their pipeline.
Capital Edge publishes one note a month on UK bridging finance, paid acquisition, and AI-led outbound. Written for brokers, by the team running the playbook.